marıo 🌊

[vjeux ✪ @Vjeux]

Custom CSS = You Lose

10 years of working on front-end codebases at Meta taught me that if you have a good component library, you should be able to just compose them to build all your screens.

All the places that use custom CSS tend to be liabilities. They won't work well in dark mode, will have weird spacing after the next design update, will look broken when embedded in other apps.

The usual excuse is that the specific product is a snowflake and badly needs it. But the reality is that most people are not doing anything novel. So it's either a failing of the builders for not using the system or the system not providing the right abstractions.

We applied these learnings with Astryx. If you see it litter your code with custom styles this means that it failed. We tried really hard to bend the agents this way but they have been heavily fined tuned to do it.

The upside is that if you're custom CSS free, most of the pages in your app/website will look consistent, have less bugs and let you update theme project wise without too much hassle.

[Andy Pavlo (@andypavlo.bsky.social) @andy_pavlo]

👊I've joined the crew at @ClickHouseDB to establish the *ClickHouse Labs* research group. We'll be working side-by-side with ClickHouse devs to cook up new ways to make the system even faster. I'll also be exploring new topics related to PostgreSQL. https://t.co/7VFLQzAL8g

[Apple TV @AppleTV]

See you on the other side, cowboy.

#Neuromancer — January 22

[Dominic Gannaway @trueadm]

Happy to announce Octane – the successor to Inferno.

React’s programming model, compiled.

Hooks, Suspense and Actions, compiled ahead of time. No virtual DOM. No Rules of Hooks. No manually maintained dependency arrays. The compiler understands what your code captures and generates the reactivity for you.

Octane supports TSX and JSX, along with TSRX for even better DX and performance.

It includes integrations for libraries such as TanStack, React Three Fiber, Lynx and StyleX, plus a compatibility layer for using Octane components inside existing React apps.

Octane is available in alpha today, with beta coming soon.

[Kimi.ai @Kimi_Moonshot]

Releasing the model weights and technical report of Kimi K3.

Kimi K3 is our most capable model: a 2.8T MoE model with native visual understanding and a 1M-token context window.

New model architecture: 2.5x the intelligence per unit of compute, not just more params.

Alongside Kimi K3, we're opening up more of the stack behind it — high-performance attention kernels, MoE communication library, and infrastructure for running agent environments at scale.

Model weights: https://t.co/7m7eEg6Y0B
Tech report: https://t.co/yeu6cjpMCT
Tech blog: https://t.co/YTfiMSNM1f

[Cult.Repo @CultRepo]

Watch the premiere of 'The Java Story' live on CultRepo on July 17th at 7pm UTC. James Gosling and other Java legends will join in the chat.

https://t.co/cBEi5tgCBf

Miss the premiere? The full documentary drops on demand right after.

@java @IBM @jetbrains @AzulSystems

[Marvel Studios @MarvelStudios]

Doomsday arrives December 18th.

Experience Avengers: Doomsday on bigger, brighter, & more immersive screens. Tickets are now on sale exclusively for Infinity Vision certified theaters: https://t.co/2HHPv9pj8t

[Mitchell Hashimoto @mitchellh]

https://t.co/RZlp39Wkrf

[Kimi.ai @Kimi_Moonshot]

Introducing Kimi K3: Open Frontier Intelligence

🔹 2.8 Trillion Parameters, 1 Million Context, Native Multimodal
🔹 Kimi Delta Attention enables up to 6.3x faster decoding in million-token contexts
🔹 Attention Residuals deliver ~25% higher training efficiency at <2% additional cost
🔹 Built for long-horizon agentic coding and self-evolving workflows

Kimi K3 is now live on on https://t.co/zrk6zZxZUo, Kimi Work, Kimi Code, and the Kimi API.
Open Weights by July 27, 2026.

🔗 API: https://t.co/XCrgjXAqMw
🔗 Tech blog: https://t.co/YTfiMSNM1f

[TypeScript @typescript]

📣 The moment is here. 📣

TypeScript 7 is officially released! 7️⃣

https://t.co/D7Mbq7vWpy

[Astryx from Meta @Astryxdesign]

🚀 Introducing Astryx from Meta – an open source design system that’s AI fluent and fully customizable without dependencies. 150+ accessible components, brand-level theming, ship-ready templates, and a CLI. Built at Meta over 8 years, now yours. Astryx Design → https://t.co/rBHFBYS76h and GitHub → https://t.co/OyHbxy4Jrl

[marıo 🌊 @mrmntte]

@LaSalitadelSin @Sinverguenzas_P jajaj "como el zorro" xd

[Alfonso López-Chau @LopezChauNava]

Hoy 27 de Junio del 2026 me dirijo a todos los militantes de AHORA NACIÓN y a todas las mujeres y hombres del Perú para expresar mi más firme solidaridad con el periodista Gustavo Gorriti.

Gustavo Gorriti es objeto de un abusivo ataque y persecución por el único "delito" de luchar contra la corrupción, por la democracia plena y por el equilibrio de poderes.

Gustavo Gorriti es un periodista independiente, es una pluma cargada de principios y de ideales por la justicia y la libertad. Es uno de los valerosos 9 millones de peruanos que se expresaron y luchan por la recuperación moral del Estado.

#AlfonsoLópezChau #AhoraNación #ConstruyendoNación

[Ihtesham Ali @ihteshamali]

China just released an open source AI model that matches the best closed models from OpenAI and Anthropic. Gavin Baker explained exactly how they did it and the answer should concern every American AI lab.

The model is called GLM 5.2. It was built by Z. AI.

You get 744 billion parameters, 1 million token context window and its MIT license, meaning anyone can download it, fork it, build a company on it, with no restrictions and no Dario.

It scored 51 points on the artificial analysis intelligence index. The highest score any open weight model has ever achieved.

It beat GPT 5.5 on the frontier software engineering benchmark. It trails Claude Opus 4.8 by less than one percentage point. And it costs 85% less to run than GPT 5.5 for comparable performance.

Gavin Baker said on the All-In podcast that this model has challenged some of his beliefs. Then he explained how China built it.

The method is called distillation.

Just think of tens of thousands of phones and computers running simultaneously, all hitting the frontier model APIs through masked accounts, asking specific questions, and harvesting what happens inside the model when it answers. Every reasoning step, every token. The entire thinking process gets recorded and fed back into the Chinese model during training.

It is a cheat sheet. It is the answer key to the exam.

And here is the part that should worry everyone.

Sacks said it plainly. China was already nine months behind American models. But now that GLM 5.2 is good enough to run its own reinforcement learning, it can improve itself without needing to distill from American models anymore. The cheat sheet let them get close enough to start writing their own answers.

Sacks said we are six months behind on the model and 24 months behind on silicon and they are only a few months behind in total.

The Z. AI founder told Elon Musk directly that open weight fable-level capability will be here before Q1 2027.

Every restriction Anthropic lobbied for, every self-imposed safety guardrail, every month of delay in releasing American frontier models accelerated this. The Chinese labs were not under those restrictions. They were not going to wait.

The composable model future Gavin described, where every enterprise runs a frontier model alongside their own fine-tuned open weight model, is coming regardless of what American labs do next.

The question is just whether the open weight half of that stack is American or Chinese.

Right now it is Chinese.

WATCH THE FULL PODCAST ON @theallinpod

[marıo 🌊 @mrmntte]

https://t.co/sNioyuc6ZZ

[lauren @poteto]

https://t.co/6Ta0f4J3u2

[Netflix @netflix]

Here's your first look at THE ONE PIECE, WIT Studio's all-new anime adaptation, coming February 2027.

[marıo 🌊 @mrmntte]

https://t.co/kO0yBmmYq9

[Carlos Viguria @cviguria]

No sé quiénes son, pero me dieron risa…

[Julia Elena 🇵🇪 @juliaelena23]

Porfa, no etiqueten a Alberto de Belaunde, RMP, Juliana Oxenford y otros periodistas que gritaban a los 4 vientos su voto nulo viciado y que ahora están aterrados con los resultados a favor del Fujimorismo.
Q tal concha!!